Projects

Outside of class, I work on various CS projects to sharpen my skills and learn new concepts. I will provide a short overview of my projects here. If you want more detail or to see the code for these projects, check out my GitHub.

LexForum

As the final project of the CodePath WEB102 Intermediate Web Development course, I created a web application called LexForum using React and Supabase. It is a forum where users can post announcements, ask questions, and comment on other posts regarding everything related to Lexington, Massachusetts. The application allows for user authentication, filtering posts by keywords, and editing and deleting posts that only the user has created.

The site is hosted using Netlify here, and the code can be found here.

LexForum Demo

AI Attendance System

I participated in the Fall 2024 AI4ALL Ignite Accelerator program. In this program, I learned about numerous AI concepts, best practices, and ethical concerns of developing AI technologies. We were put into groups, and each group created an AI-powered project based on a particular subfield. Our group focused on computer vision, so we developed an automated smart attendance system that uses facial recognition to take attendance and logs each entry's attendance. The user is also able to look at past attendance for any particular day.

See more details here.

AI Attendance Demo

Car Driving Reinforcement Learning Model

I created an RL Model that learns to drive a car along a circular track. I created a simple game as the base environment in which the player uses the arrow keys to control the car. The objective is to complete a lap as fast as possible, and if the car goes off the track, it resets at its starting position. I utilized an Epsilon-Greedy Q-Learning algorithm to implement the RL model. After about 60,000 iterations, the car was able to drive a full lap on its own.

See more details here.

RL Model Training

RL Model Full Lap

Top Drives Rating Predictor

For a while, I played a mobile game called Top Drives in which cars are given ratings based on a number of factors including power, weight, drivetrain, tire type, and more. I created a neural network trained on data of 435 cars in the game to predict the rating of new cars given stats entered by the user.

See more details here.

Top Drives Model Test